Back to Analysis

About OmicsDiscovery

OmicsDiscovery is a web-based workbench that lets cancer researchers build, explore, and analyze transcriptomic atlases from cancer cell lines — without writing a single line of R code. Search across thousands of cell lines, run differential expression, visualize results with publication-ready plots, and download everything in one click.

What can you do with it?

  • Search and select cancer cell lines from the built-in BC-84 atlas, DepMap, or GEO
  • Upload your own CSV/Excel expression matrix and analyze it the same way
  • Run MA plots, volcano plots, heatmaps, PCA, and normalization with custom thresholds
  • Perform GO/pathway enrichment across 10 databases including GSEA and PPI networks
  • Export publication-ready PNG plots, CSV tables, and multi-sheet Excel atlases

Features

Search Cell Lines

Search across 70+ breast cancer cell lines in the built-in BC-84 atlas. Find cells by name (e.g. MCF7, T47D), subtype, or tissue of origin.

DepMap Integration

Search ~2,000 cancer cell lines from the Cancer Dependency Map. Filter by tissue type (lung, breast, blood, etc.). Fetch expression TPM data for any cell line.

NCBI GEO Fetch

Import any public RNA-seq dataset from NCBI GEO by accession number (e.g. GSE48216). Auto-detects log-scale data and converts to linear for consistent analysis.

Upload Your Data

Import CSV, XLSX, TSV, or TXT expression matrices. The tool auto-detects gene columns and cell line samples. Your data is normalized into a unified 4-sheet atlas format.

MA Plots

Visualize log₂ fold-change vs. mean expression. Blue = up-regulated, red = down-regulated, grey = not significant. Configurable log₂FC and adjusted p-value thresholds.

Volcano Plots

Plot statistical significance (-log₁₀ p-value) vs. magnitude of change (log₂FC). Quickly identify the most differentially expressed genes at a glance.

Heatmaps

Cluster and visualize expression patterns across any number of cell lines. Uses ComplexHeatmap with variance-based top gene selection and z-score normalization.

PCA Analysis

Principal Component Analysis with variance explained per component. Visualize sample clustering and identify outliers in your dataset.

GO & Pathway Enrichment

Comprehensive enrichment across 10 databases: GO (BP/MF/CC), KEGG, Reactome, WikiPathways, Disease Ontology, MSigDB, ChEA, KEA, plus GSEA and PPI network analysis.

Normalization Pipeline

11 configurable knobs: scale detection, deduplication, KNN imputation, gene/sample filtering, batch correction (ComBat), CPM+log₂ transform, DESeq2 size factors.

Export Everything

Download publication-ready PNG plots, CSV expression matrices, and multi-sheet Excel atlases (Expression_Matrix + Metadata + Gene_Annotations + Summary).

Reproducible Sessions

Save your complete session (selected cells, thresholds, results) to your browser. Load it later to resume exactly where you left off.

How to Use

Four simple steps from data to publication-ready results.

1

Choose Your Data Source

Use the built-in BC-84 breast cancer atlas (70 cell lines), search DepMap for ~2,000 cancer cell lines, fetch any RNA-seq dataset from NCBI GEO, or upload your own expression file.

2

Select Cell Lines

Click cell cards to select them. You can select as many as you need. Use "Select All" to select all available cells at once. Fetched cells are automatically selected for you.

3

Configure Analysis Parameters

Adjust log₂ fold-change threshold, adjusted p-value cutoff, and number of top genes using the sliders. Set your reference/control cell line (defaults to MCF10A).

4

Run & Download Results

Click any analysis button (MA, Volcano, Heatmap, PCA, GO, Normalize). Results appear in tabs with plots, summary statistics, and download buttons for PNG, CSV, and Excel.

Enrichment Databases (12 analyses)

The GO enrichment tool runs over-representation analysis (ORA) and gene set enrichment analysis (GSEA) across these databases simultaneously:

GO BP
GO MF
GO CC
KEGG
Reactome
WP
DO
MSigDB
ChEA
KEA
GSEA
STRINGdb

Supported File Formats

.csv.xlsx.tsv.txt

Files should have genes in rows and cell lines/samples in columns.

The first column should contain gene IDs or gene symbols (e.g. ENSG00000141510 or TP53).

Expression values can be raw counts, TPM, FPKM, or log₂-transformed — the tool auto-detects and converts.

Powered by Peer-Reviewed Science

All analyses use established R/Bioconductor packages that have been published in scientific journals:

DESeq2

Love et al. (2014) Genome Biology

clusterProfiler

Yu et al. (2012) OMICS

ComplexHeatmap

Gu et al. (2016) Bioinformatics

STRINGdb

Szklarczyk et al. (2023) NAR

ReactomePA

Yu & He (2016) Mol. BioSystems

GEOquery

Davis & Meltzer (2007) Bioinformatics

OmicsDiscovery · Interactive transcriptomic atlas builder & explorer

TypeScript owns the product; R owns the science.